# Teacher support and L2 writing engagement in GenAI-integrated classrooms: a serial mediation model of growth mindset and needs satisfaction

## TL;DR

This study explores how teacher support influences student engagement in L2 writing when using AI tools, showing that growth mindset and needs satisfaction play key roles.

## Contribution

The study introduces a serial mediation model linking teacher support, growth mindset, and needs satisfaction in GenAI-integrated L2 writing classrooms.

## Key findings

- Teacher support positively affects needs satisfaction and growth mindset, which in turn boost L2 writing engagement.
- Competence support influences engagement through both direct and indirect pathways involving growth mindset and needs satisfaction.
- Emotional and autonomy support are directly and indirectly linked to engagement via needs satisfaction.

## Abstract

Teacher support is a well-established predictor of L2 writing engagement. However, the mechanisms through which it operates remain insufficiently understood, particularly in classrooms where Generative AI (GenAI) tools are integrated into instruction. Drawing on Self-determination theory (SDT) and Control–value theory (CVT), this study conceptualizes teacher support as the environmental input, growth mindset as a control appraisal, and needs satisfaction as the motivational state closely linked to engagement.

Survey data were collected from 366 Chinese university EFL learners enrolled in GenAI-integrated L2 writing classrooms and analyzed using structural equation modeling.

Results showed that all three dimensions of teacher support and growth mindset were positively associated with needs satisfaction, and that both growth mindset and needs satisfaction positively predicted L2 writing engagement. Competence support was linked to engagement through separate pathways involving needs satisfaction and growth mindset, as well as through a sequential growth-mindset → needs-satisfaction mechanism. Emotional and autonomy support showed significant direct associations with engagement and indirect effects via needs satisfaction.

These findings provide initial evidence for an SDT–CVT–aligned process, in which higher levels of teacher support are associated with higher engagement through appraisal-driven motivational states within GenAI-integrated instructional contexts. This offers practical guidance for sustaining learner engagement in GenAI-integrated L2 writing.